Although several popular theories postulate that the capacity to engage in multisensory integration is already present at birth, and is largely insensitive to experience, observations from the past grant period suggest quite the opposite. Using midbrain and cortical multisensory neurons as models, we found that this capacity develops only gradually during postnatal life and requires sensory experience for the maturation of its underlying neural circuit. Indeed, the superior colliculus (SC) model from which most of the data were collected also showed that the nature of early cross-modal experiences drive the development of this capacity and determine its form in adulthood. Using this model, the present proposal will identify the fundamental epigenetic principles that guide the postnatal development and elaboration of multisensory integration (i.e., the developmental foundations for solving the multisensory binding problem). Our hypothesis is that the solution lies in the brain's use of a statistical approach by which general operational principles are extracted from the host of early life experiences with cross-modal events to govern how neurons synthesize information from different senses and effect overt behavior. In this way, multisensory integration is crafted to function optimally in the specific environment in which it will be used. Its dependence on the covariance statistics of cross-modal stimuli establishes the maturation of multisensory integration as fundamentally different from that of its modality-specific counterparts; not only in its experiential requirements and the strategies used to encode this information, but also in its maturational time course. This is evident even in the same neurons, as multisensory neurons process both unisensory and multisensory information. Furthermore, we hypothesize, that in the case of multisensory integration, the instantiation of this guiding experience is via the cortico-SC circuit, which is the substrate on which different cross-modal experiences are superimposed, competing for representation based on their statistical likelihoods. And finally, although we believe that the effects of experience on integrative strategies to be greatest during early life when the brain is most plastic, the preset proposal will also address the possibility that these strategies can also be utilized later in life given the appropriate circumstances. PUBLIC HEALTH RELEVANCE Although the approaches used here target fundamental neural features of multisensory integration, the theoretical framework bears directly on issues relevant to a host of maturational dysfunctions in sensory processing, including those manifested in ADHD, dyslexia, autism, and several trauma-induced developmental disorders. For example, it is quite possible that the various components of multisensory integration (i.e., latency shifts, changes in response magnitude, or changes in the information rate) have different developmental time courses, sensitivity to experience and utility at different life stages. If so, there will likely be considerable differences in the consequences of interrupting their elaboration at any of those stages. This information holds promise not only for explaining some of the striking variations in the dysfunction profiles of patients with early sensory processing disorders, but also for insights into their etiology and rehabilitative strategies for dealing with these anomalies. Such information will also be of more than passing interest to patients born with visual and/or auditory deficits that are later corrected via drug therapy or prosthetic devices, and who may have to integrate some cross-modal inputs for the first time.